(Mis)adventures in trying to promote a book on AdWords | Hacker News Reader

(Mis)adventures in trying to promote a book on AdWords

trevorburnham.posterous.com

58 pointsTrevorBurnham21 comments

Open articleView on HN

Here are a few tips that I always find useful for any AdWords campaign I setup.

* Select Google Search only

* Geo-target to US/Canada only (or add more countries if relevant)

* Enter negative keywords for the campaign that we know aren't buyers for the product (free, hack, crack, etc)

* Bid on Exact Terms and Broad Matches separately at different prices

* Track all keywords through analytics to see what the bounce rates and PV/Visit are to see if the traffic is bad or the site is failing to convert the visitors.

Google AdWords still work well for a number of campaigns I'm managing in different vertical industries. It does take time and effort to set them up right through, it's not as easy as just throwing in some keywords and waiting for cash to roll in.

Thanks, these are good tips. Could you point me to a resource on setting up Analytics to track individual keywords? Right now all referrals from AdWords look the same to me ("referring site: googleads.g.doubleclick.net").

You need to link your Google Analytics account to your Adwords account and all the information will flow automatically. After doing this, in your GA account under Traffic Source, the AdWords will populate with all your stats.

http://www.google.com/support/analytics/bin/answer.py?answer...

As far breaking down the Partner Sites performance, since that's usually the first thing I turn off... if I remember correctly GA will bunch all the sites under (content) word. Perhaps someone else here has a quick answer for you.

So, whenever you see "Quality Score", I want you to read BIG Score: Best Interests of Google. This will remind you that QS has nothing to do with quality and is designed to maximize revenue for the borg. For example, the dominant factor of QS is not relevance, but is, in fact, click-through rate. (Google defines that which results in you paying more money to them as "relevant.") Next up is account history, for much the same reason that e.g. an airline will throw you off the plane if Mr. Super-Diamond-Elite really wants your seat.

Defaults are also set to be notoriously BIG. For example, no responsible AdWords professional would suggest putting search and content ads in the same campaign, because the performance of the two is not comparable and this will result in massive overspending on the Content Network. Guess what the default behavior is? Yep, both checked. Default ad rotation strategy? BIG (n.b. which is sometimes in the best interest of the advertiser, too, if they're doing it wrong). The behavior of various match types? BIG etc, etc

Sure. I don't imagine Google to be a public utility. But I am surprised that Google thinks it's in their best interests to have zero advertisers for the keyword "CoffeeScript" rather than one (clearly relevant) one.

no responsible AdWords professional would suggest putting search and content ads in the same campaign, because the performance of the two is not comparable

What is the difference in performance?

We haven't yet started a paid advertising strategy and are thinking of beginning one in the next few months. Google sent us a gift card to kick start a small (very small) campaign and we're debating the best way to use it.

The difference is going to vary depending your situation (product, keywords, ads, etc) - the key part is to run your ads separately so that you can monitor the difference.

PS> If your gift card isn't for $100 then may want to do some hunting around.

The price for content is much cheaper , if you going to have the same bid for both networks you going to spend huge amount of money for nothing.

The explanation is too much anti-Google. You are saying that QS is to maximize google's revenue ,so why is that when you have high QS, you basically paying nothing for the bids? Google punishes badly structured accounts that don't know how to advertise , takes awhile to learn what really works but it's really possible to get scores in 8-10 range.

I don't really know why the defaults are so large , but for new user I would assume is better to cast your net wide and then filter out what does not work for you. Yes , it cost money in the beginning.

Google's revenues are completely dependent on click-fraud traffic from their "partner sites" and fraudsters.

Nice of them to have this: http://www.google.com/domainpark/

In your post you claimed 75% click fraud (about 3x over legit traffic).

In my own experience it was 30% link-farm traffic about 3 years ago.

And I was being very modest with that figure so not to sound like I was making this up.

Whoa, whoa, let's be clear—I wasn't claiming 75% click fraud (though I'm certain some click fraud is occurring). I was just saying that 75% of my ad impressions are on Google's partners rather than on Google Search.

But you are claiming that those "partner sites" (that you viewed) are completely shady. Trust me, link farm and auto-generated pages are click fraud pages. Some directly, some indirectly. The effect is the same.

I think Trevor was claiming some of those partner sites are / may be completely shady.

Google adsense has stopped being relevant a long time ago... Specially for technical subjects, since most tech people just blog these ads anyway.

Moreover, to be successful with Google adwords, one needs to have a lot of money to tweak the campaigns. The system created by Google is so complicated, that only spammers and big corporations have enough money to win the game.

So what ad system is relevant?

Blast from the past: hand picking websites with the right audience for your product and advertising there with a creative tailored to the site.

BuySellAds.com helps with finding sites to advertise on directly, but most of the sites listed there are clustered around a small number of topics.

Or, if the option is there, make a killer product and attach a good affiliate program to it. I've long been amazed Pragmatic doesn't have one. I have the perfect audiences for much of their stuff but there's not much motivation to rave about it and, even when I do, I might as well link to Amazon (meaning they make less money).

I linked up Trevor's book in JavaScript Weekl recently simply because it was such a big deal but I couldn't make a bean from it (not always the point, but if I have two choices..)

Hey, thanks for that! I'll prod Andy/Dave about why we don't have an affiliate program.

With these sorts of stories, it makes me wonder if Google's relevance could drop in the future. Success through the system seems to be gamed.

"In my last post, I described my efforts to get my book, CoffeeScript: Accelerated JavaScript Development, near the top of the Google results for the search term coffeescript. Since that post went up, it seems to have gone from #10 to #8, so that's progress. Or it could just be a random fluctuation. Google has been known to show search results in different order to different people on different devices at different times, after all."

- email me (mail's in profile), if you want a free inboxSEO account to track your rankings for this project... i would have posted this on the blog but there's no way to comment there...

Trevor if you need help, feel free to send me a msg via email. (in my profile) We can get Google Analytics working properly for you and that will help you maximize your adwords spending for ROI. If you have direct access to your sites code (I assume yes from your book) we can set you up in a few minutes and get you going. (The help offer is gratis not paid consulting)

All you should need is a basic GA setup and a basic adwords account to start making sales.

Reply on news.ycombinator.com